The inevitable 1.9 and what comes after

[ZooNamedGames]

We've finally reached 1.6.1. An achievement for KSP's development. As someone who's been around since 0.19 in 2013, it's been one hell of a ride. But as someone who's watched and observed the game's development over the years, I am confident in expecting a 1.7, and the unavoidable 1.7.1 that comes after it. With 1.8 and 1.8.1 to follow after it, but it forces a greater question in the end; what's coming after 1.9? The obvious answer would be 2.0, but with my experience of the 1.0 update, and more importantly, the likely future of KSP, could 2.0 be Squad's first release of Kerbal Space Program 2? Or will 2.0 just be another update and we continue past it without incident? Or, will we shift to 1.9.1, 1.9.2, 1.9.3, and so on?

Though this seems like a rather minute issue, I'm merely curious as to the development plan for KSP. A question I know the KSP developer's have been hesistant to answer as to avoid having to make promises they can't uphold or will have to later break/fall short on due to changing circumstances, but at the same time, simply being aware of what's on the table, is a lot easier to prepare for as a player and more importantly, as a paying consumer.

I've been on this ride for a long 6 years now. I'd like to know what the plan might be for year 7. I'm curious what everyone else thinks though.

^{_{Also, sorry for posting two posts in the same section back to back, I just felt both were relevant and worthy of discussion. So please forgive me.}}

[Superfluous J]

Software versioning is not math. 1.9 plus 0.1 is not 2.0, but instead is 1.10.

[MaverickSawyer]

Yep, It'd be 1.10., just like how the development versions of KSP went through 0.10.x without going into full release, aka 1.0. And you even mentioned having been around since 0.19, so you know that it'd go to 1.10, wouldn't you? After all... how else would you get to 0.19?  

[klgraham1013]

6 hours ago, 5thHorseman said:

Software versioning is not math. 1.9 plus 0.1 is not 2.0, but instead is 1.10.

Probably this.

...but who even knows.  The money's going to run out eventually.  I just pray we're not stuck with a version that has a nasty new bug that wasn't able to get fixed.  Hopefully they see the end coming and can prepare with a few bug fix only releases.  I'd hate for KSP to have the carpet pulled out from under it.

7 hours ago, ZooNamedGames said:

Kerbal Space Program 2

I really can't wait for this, honestly.  A game built from the ground up, knowing what it's going to be, and what the visuals should look like, and higher minimum specs.  No more talk of potatoes and they're inability to run X feature so it shouldn't be in the game.  It could be brilliant.

Edited January 15, 2019 by klgraham1013

[steve_v]

7 minutes ago, klgraham1013 said:

I just pray we're not stuck with a version that has a nasty new bug that wasn't able to get fixed.

So do I. But you know, learning from history and all that.
I mean, joystick support has been broken since 1.4, wheel physics has been borked in a variety of ways since 1.1, and every new release seems to include some new bug in core systems. Drifting orbits, misaligned terrain tiles, terrain collision detection problems, wheel physics every damn release, the recent part database screwup, ongoing garbage collection related performance problems.

Maybe I'm just pessimistic, but I've been waiting several years now for the 1.0 stable release, and it hasn't happened. 1.3.1 was close, but no cigar...
Then 1.4 came around, and instead of fixing the terrain collision detection on scene load, they broke joystick support. What the actual fark?
Neither problem fixed as of 1.6.1, surprised anyone? Perhaps I should wait for 1.9? Unless an engine update screws something else before then...

I'd just love to see a stable release with no serious bugs, then I'd be happy to call it a day and maybe buy some DLC or something.
I don't care about new parts or texture updates. Fix the bugs and stick a damn fork in it already. 1.7 would be a fine time.

[Superfluous J]

Thread with question about software versioning turns into Squad hate fest. Check.

[ZooNamedGames]

1 hour ago, 5thHorseman said:

Thread with question about software versioning turns into Squad hate fest. Check.

I don’t see the hate, merely some concerns after discussing the topic at hand.

[razark]

8 hours ago, 5thHorseman said:

Software versioning is not math. 1.9 plus 0.1 is not 2.0, but instead is 1.10.

I'm not so sure.

0.25 + 0.1 = 0.90?

[steve_v]

1 hour ago, 5thHorseman said:

Thread with question about software versioning turns into Squad always buggy releases hate fest.

FTFY.

2 hours ago, klgraham1013 said:

A game built from the ground up, knowing what it's going to be, and what the visuals should look like, and higher minimum specs.

So long as it's not built on Unity3D, bring it on.

[klgraham1013]

6 minutes ago, steve_v said:

So long as it's not built on Unity3D, bring it on.

I'm no expert, but is any engine actually suited for KSP?  I've never really been in the #BlameUnity camp.

[IncongruousGoat]

3 minutes ago, klgraham1013 said:

I'm no expert, but is any engine actually suited for KSP?  I've never really been in the #BlameUnity camp.

*puts on programmer hat*

Not really, no. The problem with KSP is that it has to compute accurate physics at enormous distances, beyond the capabilities of even double-precision floating-point arithmetic. This isn't something any engine is really designed to do, because very, very few games have a game world as big as KSP's and so don't ever run into floating-point precision issues. A custom engine using quadruple-precision floating point wouldn't help either, since the x86 instruction set doesn't natively support quadruple-precision floating point, or any 128-bit arithmetic for that matter. It's possible to get around this using an emulation library, but these are slow.

That said, a good few of the notable bugs that have cropped up over the years (looking at you, wheels) have been due to problems with Unity itself, independent of the stuff that makes KSP not good for any engine. So a new engine would definitely help with at least some bugs, although it would almost certainly come with bugs of its own. It's hard to abuse an engine as hard as KSP has to and not have strange things happen.

[Superfluous J]

1 hour ago, razark said:

I'm not so sure.

0.25 + 0.1 = 0.90?

Like I said, not math.

[steve_v]

46 minutes ago, klgraham1013 said:

I've never really been in the #BlameUnity camp.

I'm in the blame unity (and those who pick buggy Unity releases) "camp" for things that are unity problems.
Wheels spring to mind, as does joystick support going AWOL, various longstanding physics bugs, the release that crashed constantly with memory-management errors, the release that crashed initialising audio, and the release that crashed people's window managers. That's in no way a complete list.
My primary problem with Unity is that every engine update seems to bring new problems, especially for the GNU/Linux and MacOS builds, and that Squad refuses to do anything about them.
Perhaps this is partially due to Unity not cooperating, but regardless of who is to blame engine bugs consistently remain unfixed for entire release cycles.
 

40 minutes ago, IncongruousGoat said:

Not really, no.

Indeed. I agree that what KSP is doing is a tall order for any game engine, particularly at the scales involved.

Though you may note that my complaints wrt unity, with the possible exception of the orbit drift (which was a regression anyway) aren't related to the limitations of floating-point math.
They are all either deficiencies in basic game engine - OS interfacing or in mechanisms provided by Unity - such as the wheel system, the runtime garbage collector, or the memory chewing UI.

1 hour ago, razark said:

I'm not so sure.

0.25 + 0.1 = 0.90?

Do you remember saying it wasn't ready to go beta, or that the single beta wasn't enough? I do.

[razark]

3 minutes ago, steve_v said:

Do you remember saying it wasn't ready to go beta, or that the single beta wasn't enough? I do.

I don't remember much of an issue when they announced it was going beta, and said they'd have a cycle of testing/release/testing/release/etc.

I do recall a massive surge of opposition when they announced the very next release would be 1.0, including new, untested features.

[steve_v]

1 minute ago, razark said:

said they'd have a cycle of testing/release/testing/release/etc

+

1 minute ago, razark said:

the very next release would be 1.0, including new, untested features.

Is what got my goat. And we were right too, 1.0 aero was a shambles and the new thermal system didn't get fixed until 1.1 - which crashed so much it was unplayable.*

*I think I have my releases right, it's been a while. Correct me if not.

[klgraham1013]

Am I wrong in thinking there must be a way to get around Unity's issue?  Do you absolutely need to use Unity's built in code? Can you purchase a license to the source code, and modify the code to your games needs?  Are we really so trapped behind engines in game development?

Also, I'm under the assumption that Squad isn't forced to change the version of Unity it's on.  Surely Squad saw incoming problems while developing on the newer version.  Could they not stop and go back?

Edited January 15, 2019 by klgraham1013

[steve_v]

57 minutes ago, klgraham1013 said:

Am I wrong in thinking there must be a way to get around Unity's issue?

There is always a way. Whether there's a sensible or economical way is another matter.
 

57 minutes ago, klgraham1013 said:

Do you absolutely need to use Unity's built in code?

For many things, no. You can write your own as plugins, it's just a bunch of work.

The possibility of replacing unity's wheel code has been raised before, and a plugin that might do the job suggested.
The response was, IIRC, "That's not something that can be bought in the unity store and used right now,  getting it to work with KSP would be a lot of work."

I can't help wondering if developing a custom wheel plugin would have taken 2.5 years, because that's about how long they've been fiddling with the one unity provides, and it's still nearly as awful as the day it landed.
 

57 minutes ago, klgraham1013 said:

Can you purchase a license to the source code, and modify the code to your games needs?

Sure, unity source-code licences were available last time I checked. Game developers modify engines all the time, and they've been doing it since IDtech 1 and Heretic.

Purchasing one was suggested repeatedly in discussion of the "double-free or corruption" crashes 1.1 suffered, the licence was deemed too expensive and debugging unity's code too difficult.

 

57 minutes ago, klgraham1013 said:

Surely Squad saw incoming problems while developing on the newer version.

One would think so.

My guess is that they either do not test sufficiently before committing to a target version, or new unity releases have brought things they absolutely wanted for a new feature and someone decided it was a reasonable tradeoff.

In the case of the update that came with 1.4 and borked joystick support on GNU/Linux, I suspect it was the latter, and that the feature was something needed for the DLC.
This suspicion is one reason this particular issue annoys me as much as it does.

Edited January 15, 2019 by steve_v

[Superfluous J]

It is also possible that one or more of the people who no longer work for Squad made the decision, and that decision was why they no longer work there.

Note the above is pure speculation and not based on any facts. However it's important to note that Squad is not a person but a group, that has changed quite significantly over the years.

[klgraham1013]

1 hour ago, 5thHorseman said:

Squad is not a person but a group, that has changed quite significantly over the years.

Indeed it has.  These days, we barely have any idea of who works at Squad, and those we do know, we have little idea what they actually do, or what they worked on.  Times have definitely changed.

[Superfluous J]

33 minutes ago, klgraham1013 said:

Indeed it has.  These days, we barely have any idea of who works at Squad, and those we do know, we have little idea what they actually do, or what they worked on.  Times have definitely changed.

Well if my fanbase was so acidic I'd probably not bother either.

[klgraham1013]

6 hours ago, 5thHorseman said:

Well if my fanbase was so acidic I'd probably not bother either.

This is the least acidic fanbase I've been around.

[Superfluous J]

46 minutes ago, klgraham1013 said:

This is the least acidic fanbase I've been around.

It's about typical for the ones I frequent, except FTL that's basically sunshine and roses. I expect AAA game forums are ridiculously worse, but I tend to prefer indie games.

One thing is consistent, though: All of the communities had greater developer interaction before the incessant complaining started.

[steve_v]

5 minutes ago, 5thHorseman said:

I expect AAA game forums are ridiculously worse

Personally, my expectations of a products quality scale fairly linearly with the size and potential funding of the company producing it.
Hence, I expect better from KSP now that it's owned by TTI, and likewise my expectations increased when it went 1.0 on steam etc. at a non-indie price.

 

9 minutes ago, 5thHorseman said:

All of the communities had greater developer interaction before the incessant complaining started.

All the communities I have been involved in had less complaining before the lack of developer interaction and incessant pr-speak started. Which came first?

[flart]

18 hours ago, razark said:

0.25 + 0.1 = 0.90?

You was joking and don't need an answer, but there it is. I usually do something like that (similar squad did):
0.1.x - Alpha
0.2.x - Beta
0.9.x - RC
1.0  - Release

 

On 1/15/2019 at 8:27 AM, ZooNamedGames said:

The obvious answer would be 2.0

90% chance 1.10 will follow after 1.9, 
7% chance they take more than 3 months for development, make a little more new features and call it 2.0 for the increasing sales
3% chance they just call it 2.0.

 

[klgraham1013]

3 hours ago, steve_v said:

All the communities I have been involved in had less complaining before the lack of developer interaction and incessant pr-speak started. Which came first?

Thumbs up.

2 hours ago, flart said:

call it 2.0 for the increasing sales

Oh dear.  Your right.  From the company that went 1.0 for speculated similar reasons, this actually makes a ton of sense.  Whether or not it would actually be deserving of the 2.0 moniker.  I think my percentage chance of Squad jumping from 1.9 to 2.0 just increased exponentially.